Thread by @ZemmourMichael

Hi, let me present below "Four levers of redistribution: the impact of tax and transfer systems", recently published in the Review of Income and Wealth , by Elvire #Guillaud, Matthew #Olckers and myself.

(free access here: rdcu.be/bgJQs )

Thread

⬇️⬇️⬇️

We propose a decomposition of the four levers of monetary redistribution used in OECD countries.

To reduce income inequality, you can use either taxes or transfers. But the redistributive impact of both depends on the interaction of the average rate and the average progressivity / targeting.
This is summarized by the following equation:

In theory, the same level of redistribution may be achieved solely by transfers, solely by taxes, or by a mix.

Also, the same level of redistribution may be achieved by a low tax rate with strong progressivity or a high tax tax and weak progressivity.

The paper maps how countries use the four levers.

We are not the first to describe how countries combine taxes and transfers for redistribution and we build on the work of a number of important contributions. (See references at the end of the thread.)

We define redistribution as the difference between market income (labor+capital+pension) and disposable income.

We use @lisdata at the household level. One of our input is to impute missing social security contributions (including employer contributions) using rates supplied by the @OECD_Stat . This is important because the share of employer contrib varies greatly across countries.

Previous studies using @lisdata overlooked the role of employer contribution; missing part of the story and biasing the comparison. See our code on GitHub if you would like to include the imputations in your own projects: github.com/matthewolckers…

We compute the Gini coefficient at market income, after transfers and after taxes. Main lesson: redistribution is sizeable in all countries but it does not change an unequal country into an equal one. Market inequality = excellent predictor of the disposable income inequality

(You sometimes read in the media that some very unequal country become equal thanks to tax & benefits; this is a fallacy, due to the exclusion of public pensions from Market income, which leads to count improperly all pensioners from pub pension countries as Market-income poors)

Results:
1. We find two groups: low and high redistribution countries. The respective role of taxes and transfers varies, but tax tends to dominate transfers (pension excluded). The picture is more balanced for the working-age population.

2. Transfer redistribution = transfer size*targeting

The size of transfers is the main driver in transfer redistribution (as in previous studies). The intensity of targeting has only a second order effect.

3. Tax redistribution = progressivity * tax rate
We observe that different countries obtain the same tax redistribution with low but progressive taxes (e.g. Ireland), or with higher tax but less progressive (e.g. Sweden).

4. We observe an incompatibility between a high tax rate and strong tax progressivity.

We check our results for the subsample of working-age households and compute separately a redistributive effect of pensions.

Main caveats: 1. LIS data provides a broad comparison but might be imprecise for a given country. Our results are robust to the exclusion of any country or year.

2. We account for the taxes that impact most households but we do not measure capital gains tax, which could be important for the Top 1%. Very rich and very poor households could also be misrepresented in the survey data.

Final note: our concept of progressivity measures the share of taxation supported by different income groups, which is different from statutory marginal tax rates.

References

The research project was supported by the @LIEPP_ScPo and @EN3S and at its final stage benefited from fantastic data work by @victoramoureux & the support from the @lisdata team (@jneugsch et al. ).
It received the @lisdata Aldi Hagenaars Memorial Award as a WP.

Comments & critics welcome.
[End]